One of the most important things we as horseplayers have to do is calculate the likelihood that a particular event will take place.

On the simplest level, we need to make an odds line to predict how often Horse A will win this upcoming race.

If we have some way of doing this, we can then check the toteboard (or bookmakers in Australia) to determine whether Horse A is an underlay, about right or an overlay, and then act accordingly.

For multiple-win exotics, such as the daily double, etc., we need to multiply the chance that each single event will happen to get the overall hit possibility.

For instance, let's say we give Point Taken a 25 per cent chance to win the first leg of a double, and Tizthen a 20 per cent chance to win the second leg. The likelihood that both events will take place for a successful wager is .25 x .20, or 5 per cent. Therefore, a \$2 payoff of \$42 (also stated as 20/1) would be break-even.

This, of course, assumes that we've analysed the chances of each horse correctly.

If we play a single large ticket for a pick 3 or pick 6, this calculation isn't difficult. For instance, let's say we play a pick 3 (treble) ticket of numbers 1-4-5 into 2-3 into 6-7-9 for a total of 18 combinations.

If we figure that the combined chance of 1, 4 or 5 winning the first leg is 60 per cent, that 2 and 3 have a 50 per cent chance of winning the second leg, and 6, 7 and 9 have an 85 per cent chance of winning the third leg (these numbers always vary depending on the strength of the selections), the chance that this ticket will hit will be .60 x .50 x .85 or 25.5 per cent.

A \$1 ticket costing \$18 therefore needs to pay at least \$71 for us to break even (over a period of time).

Calculations get more tedious with multiple tickets, as in a pick 6 in which we make out 23 separate combinations. To get an accurate read on the likelihood of your victory, we'll need to calculate the win chance of each of these 23 tickets separately, then add the sums.

And, with so many odds-varied possibilities (you have a 6/5 shot and a 9/1 shot on the same ticket, for instance), it becomes extremely difficult to predict with much accuracy what such a ticket might pay if we do get lucky enough to hit it.

The volatility of such bets as pick 6's and superfectas (their low win percentage and high payoff variations) make them difficult to build a wagering programme around.

Anything that you can estimate lends itself to a calculation ... of how likely the event is to win, and what payoff you need to make it profitable. To make money at parimutuel (tote) betting, you need to be able to make these calculations. No overlay, no bet.

If you make an odds-line, the charts in my book Money Secrets At The Racetrack, show you whether a particular number is an overlay or not for daily doubles, quinellas and exactas.

Punters who simply box everything that moves run into trouble. Certain combinations are much more likely to win than others, yet these players bet the same amount on each combination with no consideration of what anything might return.

And punters who profess never to care what a horse will pay (they reckon they'll pick enough winners) are in the biggest trouble of all.

Much of the calculating we do, of course, is simply a guess. Because we're estimating how many times a horse will win out of 100 times this race will be run, and the race will be run only once, it can never be more than an estimation.

However, like the weatherman who predicts there's a 30 per cent chance of rain based on thousands of previous predictions, the longer we work at this the better our estimation should be.

The whole idea is to be able to figure whether a bet is worth playing, or should be passed.

Barry Meadow is one of America's best-known handicappers and racing writers. His monthly newsletter is available from TR Publishing at: www.trpublishing.com as are his excellent series of books on racing and harness racing, plus casino gambling.